LOS ANGELES — The World Health Organization says the H5N1 bird flu kills nearly 60 percent of people who become infected, but a study released Thursday suggests the true fatality rate may actually be much lower.
Virologists at the Mount Sinai School of Medicine in New York City examined data on blood samples collected from more than 12,000 people in Asia, Europe and Africa and found evidence of H5N1 infection in 1 percent to 2 percent of cases. Most of those people did not become ill with the flu, according to a report in the journal Science, and none of them died.
The findings, which used data from 20 previously published studies, suggest that many more people have been infected with H5N1 flu viruses than the 586 officially confirmed by the WHO as of Wednesday; if so, the fatality rate could be lower than the 59 percent reported by the global health agency.
“The World Health Organization criteria that are currently being used for confirmation of H5N1 infections are good for the identification of very severe cases, but they do not pick up the cases that are mild or asymptomatic” because such patients are less likely to seek treatment in a hospital, said postdoctoral researcher Taia Wang, who led the study.
Flu experts have speculated before that the WHO’s surveillance system is patchy and misses an unknown number of cases. There is no scientific consensus about what the true fatality rate is.
Gregory Hartl, a WHO spokesman in Geneva, acknowledged Thursday that the agency’s criteria for confirming bird flu cases might underreport mild cases. But he said it was more likely that officials were missing cases of bird flu deaths.
In some regions of the world where people have contracted H5N1, including parts of Asia, patients are unlikely to go to a hospital “until very late, if at all,” Hartl said. Anyone who died from bird flu outside of a hospital would not be included in the official tally.
Even if the fatality rate is substantially lower than 60 percent, bird flu still has the potential to pose a very serious danger to humans if it were to mutate in a way that allowed it to spread easily between people. So far, cases of direct human-to-human transmission have been rare.
“You could reduce the seriousness of this virus twentyfold and it would still exceed that of the 1918 Spanish flu,” said Michael T. Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota.
The H1N1 flu responsible for the 1918 outbreak had a fatality rate of about 2 percent. But it killed as many as 50 million people worldwide because it spread easily between people.
Flu researchers have been studying H5N1 to see if has the potential to become more contagious in people, and two research groups have created strains that pass easily between lab ferrets, which react to influenza much as humans do.
In December, fears of an H5N1 pandemic led a U.S. government biosecurity advisory board to seek restrictions on the publication of that research. The request angered scientists who said they needed to share their findings in order to develop vaccines and prepare for a possible pandemic.
But the biosecurity panel, of which Osterholm is a member, said the data could be dangerous if it wound up in the wrong hands.
Scientists and public health officials are still debating how to proceed. Last week, the flu researchers agreed to continue a moratorium on the research and further delay publication of their findings while the issues are sorted out.